Modelling and estimation of genetic parameters for milk urea content in Holstein dairy heifers

Milk urea content (MUC) routinely recorded by infrared might be used as a selection trait to prevent higher nitrogen excretion. Therefore, modelling and estimation of both heritability and genetic relationship to the important milk production traits are needed. Data of 589 heifers between the 8(th) and the 180(th) day of lactation from the bull dam performance test at the dairy research farm Karkendamm were analysed. Cows were fed a total mixed ration ad libitum twice daily. Additionally, a fixed amount of concentrates was given via feeding stations. Milk ingredients were analysed weekly, and milk yield was recorded automatically for each milking. In modelling five well established parametric functions of days in milk (ALI and SCHAEFFER, GUO and SWALVE, WILMINK, Legendre polynomials of third and fourth degree) were chosen for modelling both fixed and random regression coefficients. Model fit was judged by two information criteria and by illustration of the residuals against days in milk. Subsequently, on the basis of the best fitted random regression model variance components were estimated by REML. In addition to the fixed and random regressions the model includes the fixed effects test day and class of age of first calving. Estimations were based on univariate (heritabilities) and bivariate (genetic correlations) runs. Function of Am and SCHAEFFER seems to be most suitable for modelling both the fixed and the random regression part of the mixed model. A serious alternative is the Legendre polynomial of fourth degree. Heritability for MUC ranges between 0.28 and 0.48. The yield traits milk, fat and protein are unfavourably genetically correlated with MUC. In contrast, fat and protein content are favourably correlated with MUC. So, a consequent selection on fat and protein content could reduce nitrogen excretion.